Installation

Getting started with single-algebra is straightforward. This page guides you through the installation process and basic setup.

Adding single-algebra to Your Project

Add single-algebra to your Rust project by including it in your Cargo.toml file:

[dependencies]
single-algebra = "0.7.0"

By default, this includes the core functionality with minimal dependencies. To enable additional features, specify them explicitly:

[dependencies]
single-algebra = { version = "0.7.0", features = ["lapack", "faer"] }

See the Features page for a complete list of available features.

Basic Usage

Once installed, you can import single-algebra in your Rust code:

// Import the entire library
use single_algebra::*;

// Or import specific components
use single_algebra::sparse::MatrixSum;
use single_algebra::dimred::pca::dense::PCABuilder;
use single_algebra::similarity::CosineSimilarity;

Verifying Installation

Create a simple example to verify your installation:

use single_algebra::sparse::{MatrixSum};
use nalgebra_sparse::{CooMatrix, CsrMatrix};

fn main() -> anyhow::Result<()> {
    // Create a simple sparse matrix
    let coo = CooMatrix::try_from_triplets(
        3, 3,
        vec![0, 1, 2],
        vec![0, 1, 2],
        vec![1.0, 2.0, 3.0]
    ).unwrap();
    
    let csr: CsrMatrix<f64> = (&coo).into();
    
    // Calculate column sums
    let col_sums: Vec<f64> = csr.sum_col()?;
    println!("Column sums: {:?}", col_sums);
    
    Ok(())
}

Available Features

single-algebra provides several optional features that can be enabled based on your needs:

• lapack: Enables LAPACK integration for numerical linear algebra via nalgebra-lapack with OpenBLAS
• faer: Enables Faer for numerical linear algebra (a pure Rust implementation)
• simba: Enables SIMD acceleration for statistics
• smartcore: Integrates with the smartcore machine learning library

To enable multiple features:

[dependencies]
single-algebra = { version = "0.7.0", features = ["lapack", "faer", "simba"] }

Platform Support

single-algebra is designed to work on all platforms supported by Rust, including:

• Linux
• macOS
• Windows
• Other Unix-like systems

Certain features that depend on external libraries (like LAPACK) may have additional platform-specific requirements. See the Requirements page for details.

Release Channels

single-algebra follows semantic versioning:

• Stable releases: Use the latest version number (e.g., 0.7.0)
• Development: For the latest development version, you can use a Git dependency:

[dependencies]
single-algebra = { git = "https://github.com/SingleRust/single-algebra.git", branch = "main" }

Updating

To update to the latest version of single-algebra, modify the version number in your Cargo.toml:

[dependencies]
single-algebra = "0.7.0" # Current latest version

Or run cargo update to update within your specified version constraints.

For questions or issues with installation, please open an issue on the GitHub repository.